How ChatGPT is Transforming Trip Planning

Hong Kong, 24 February 2026 – Artificial intelligence (AI) is transforming the way we live, work – and vacation. With their rapid, sophisticated responses, AI chatbots like ChatGPT offer travellers an unprecedentedly powerful tool for trip planning. However, this tool is not without limitations, especially when it comes to narrowing down multiple options for destinations and activities. Thanks to Professor Seunghun Shin of the School of Hotel and Tourism Management (SHTM) at The Hong Kong Polytechnic University (PolyU) and four co-authors, online travel agencies now have novel insights into the impact of ChatGPT on travellers’ trip planning behaviour, specifically the relative effectiveness of ChatGPT and humans as recommendation agents.

ChatGPT, a dialogue-based AI tool that can understand natural human language and generate human-like written text, has the potential to revolutionise trip planning. “Its remarkable ability to interact with users by providing sophisticated answers to their queries has led millions of people to adopt it”, say the researchers.

Traditionally, travellers needed to comb multiple websites and other information sources – from search engines to maps, blogs and online review pages – to find the perfect destinations and activities. Today, the authors note, “they can bypass such complicated and disturbing processes through a simple conversation with ChatGPT”. The chatbot can deliver customised itineraries within just a few seconds, providing “a valid initial reference for trip planning”.

Despite these advantages, certain limitations of the tool may pose obstacles for prospective travellers. The first – at least at the time of this study – was ChatGPT’s inability to provide real-time information. Second, the tool’s focus on text rather than images or videos is inimical to the visual nature of shopping for destinations and activities. “Last but not least”, the researchers say, “ChatGPT is not always correct”.

These limitations may make travellers unwilling to fully trust ChatGPT, especially its ability to narrow down multiple valid options at later stages of the trip planning process. Unfortunately, however, research to date has paid insufficient attention to how tourists perceive the role of ChatGPT in the trip planning process. The researchers set out to fill this important gap.

Their first task was to consult the literature to develop hypotheses for testing. The most salient framework appeared to be that of the choice overload effect, or the negative consequences of a large assortment of choices. “The choice overload effect”, say the researchers, “is generally understood to mean a situation in which the intricacy of the decision problem facing a person transcends his or her cognitive capacity”.

First, having a multitude of choices can be overwhelming, due to the near-impossibility of comparing all available options. “The regret stemming from the inability to select the best choice enhances the attractiveness of other options”, the researchers warn, “potentially resulting in unfavourable outcomes such as switching decisions or opting for no decision at all”. Second, the presence of so many options raises consumers’ hope of finding the ideal choice – inevitably leading to disappointment when this becomes impossible.

In the hospitality and tourism field, researchers have sought to identify moderators capable of reducing the choice overload effect. For example, building travellers’ self-confidence has been found to mitigate the negative effects of large numbers of options at multiple stages in the decision-making process. A frequently overlooked factor, however, is the information source (e.g., AI or human), which may increase the accuracy of decision-making and thus reduce the choice overload effect.

Expanding on choice overload theory, the researchers proposed their first hypothesis. “As ChatGPT streamlines a multitude of options into a more concise selection”, they say, “it may lead to a reduction in travellers’ satisfaction with the recommendations and their intention to visit the recommended destination”. Although choice overload theory generally assumes that narrowing down options helps people make better choices, this depends on whether appealing choices are retained and unattractive ones are excluded – tasks at which ChatGPT may not excel.

The authors also proposed that ChatGPT’s perceived trustworthiness can mediate the impact of its option-reduction service on travellers’ recommendation satisfaction and visit intention. AI is widely believed to lack efficacy in handling tasks requiring intuition and empathy. “Given that the process of trip planning is subjective, intuitive, and emotional”, the researchers say, “travellers are inclined not to trust information that has been curated by ChatGPT”.

Finally, the authors identified possible boundary conditions for how travellers perceive ChatGPT when narrowing down multiple travel options. “First, we anticipate that concerns about forfeiting the best options will be more pronounced when the initial number of choices is substantial”, they say. Second, the choice overload effect varies depending on whether consumers narrow down options independently, which can minimise concerns about forfeiting the best possible alternatives. Therefore, they expected the negative effect of reducing options to decrease when this reduction is done by the human decision-maker rather than ChatGPT.

To test their hypotheses, the researchers crafted five experimental studies in which participants imagined planning trips to destinations, interacting with ChatGPT or friends to generate and narrow down travel options. Their aim was to use these studies to empirically “determine the negative effect of reducing options from the initial recommendations using ChatGPT”.

Study 1 focused on Hong Kong as the target trip destination. As hypothesised, “satisfaction with the recommendations was reduced significantly when ChatGPT suggested narrowed-down options from the initial large option”, the researchers report. In Study 2, the participants imagined planning a trip to Key West, Florida. Option reduction by ChatGPT (compared with the participants themselves) negatively affected visit intention, mediated by perceived trustworthiness and satisfaction, especially when the initial set of options was large.

Strengthening these findings, Study 3 revealed that travellers were more satisfied when they narrowed down the options themselves than when they relied on ChatGPT. In Study 4, satisfaction and visit intention were higher when options were narrowed down by human experts than by ChatGPT. Finally, Study 5 demonstrated that the negative effects of ChatGPT’s option reduction were mitigated when travellers actively engaged in the narrowing process, suggesting that user participation can buffer against the damage done to trust and satisfaction by AI-driven reduction.

“Our research reveals the underlying psychological mechanisms behind why decision-makers do not prefer a reduction in recommendations by AI agents”, the authors tell us. Their findings not only contribute to the literature but may also help online travel agencies (OTAs) more effectively embed ChatGPT in their websites to help travellers create trip itineraries.

First, their findings indicated that ChatGPT may not be helpful in narrowing down options at a late stage of trip planning, because its options-reduction ability is not perceived as trustworthy. Therefore, “OTAs should emphasise ChatGPT’s ability to generate the initial pool of destinations or activities and deemphasise its narrowing-down ability”.

“Second”, they say, “our findings showed that the benefits of using ChatGPT for trip planning can increase when its initial recommendations are narrowed down by travellers”. OTAs could thus encourage travellers to be engaged in the creation process to maximise their satisfaction with the itineraries. “Businesses could instruct their chatbots to give multiple possible alternatives to customers’ inquiries”, the authors say, “to lead them to make their own choices”.

Finally, they found that travellers perceived choice overload when more than 15 options were proposed by AI tools. This insight may be applicable beyond travel planning. “Based on our findings”, the authors conclude, “hospitality businesses could adjust the default number of recommendations their AI tools provide, to enable customers to easily interact with them, and to have better experience with such tools”.

Seunghun Shin, Jungkeun Kim, Eunji Lee, Yerin Yhee and Chulmo Koo (2025). ChatGPT for Trip Planning: The Effect of Narrowing Down Options. Journal of Travel Research, Vol. 64, No. 2, 247–266.

*****

About PolyU School of Hotel and Tourism Management

For more than four decades, the School of Hotel and Tourism Management (SHTM) of The Hong Kong Polytechnic University has refined a distinctive vision of hospitality and tourism education and become a world-leading hotel and tourism school. Ranked No. 1 in the world in the “Hospitality and Tourism Management” category in ShanghaiRanking’s Global Ranking of Academic Subjects 2025 for the ninth consecutive year; placed No. 1 globally in the “Commerce, Management, Tourism and Services” category in the University Ranking by Academic Performance in 2024/2025 for eight years in a row; rated No. 1 in the world in the “Hospitality, Leisure, Sport & Tourism” subject area by the CWUR Rankings by Subject 2017; and ranked No. 1 in Asia in the “Hospitality and Leisure Management” subject area in the QS World University Rankings by Subject 2025, the SHTM is a symbol of excellence in the field, exemplifying its motto of Leading Hospitality and Tourism.

The School is driven by the need to serve its industry and academic communities through the advancement of education and dissemination of knowledge. With a strong international team of over 90 faculty members from 21 countries and regions around the world, the SHTM offers programmes at levels ranging from undergraduate to doctoral degrees. Through Hotel ICON, the School’s groundbreaking teaching and research hotel and a vital aspect of its paradigm-shifting approach to hospitality and tourism education, the SHTM is advancing teaching, learning and research, and inspiring a new generation of passionate, pioneering professionals to take their positions as leaders in the hospitality and tourism industry.

Press contact     :   Ms Tiffany Chan, Marketing Manager

                              School of Hotel and Tourism Management

Telephone         :   (852) 3400 2293

E-mail               :   tiffany-sm.chan@polyu.edu.hk

Website             :   https://www.polyu.edu.hk/shtm/

To read the full article, please click here.